One of John Dickson Carr’s most maligned mysteries, Seeing is Believing takes a ridiculous murder plot based around hypnosis, rubber daggers and poisoned grapefruits, and then tops it off with a trick which many people think is cheating. I agree, although I don’t think anyone has yet put their finger on exactly why it’s cheating. In today’s post, I suggest that the fairness of an author’s mystery depends on a good deal of unspoken context, and that it might be possible for a trick to be fair in one author’s hands but unfair in another’s.
Seeing is Believing begins with this matter-of-fact paragraph:
One night in midsummer, at Cheltenham in Gloucestershire, Arthur Fane murdered a nineteen-year-old girl named Polly Allen. That was the admitted fact.
Only he didn’t. It isn’t a fact at all. And the only people who say it is are the murderer and Carr himself.
Seeing is Believing has an extremely elegant structure at its heart. Arthur Fane appears to be wealthy. He has a large house and all the accoutrements of fine living. He also seems to have committed a murder. He talks about it in his sleep and there are many witnesses to attest to his suspicious behaviour on the night Polly Allen disappeared. Hubert Fane appears to be poor and seems to be blackmailing Arthur – mooching off his brother’s wealth and acting like he’s master of the house.
But actually the situation is reversed: Arthur is the blackmailer. He was rich but lost all his money. He talks about the murder in his sleep because he helped conceal it, which is why he was acting suspiciously. Hubert is the murderer, and is actually rather wealthy. He doesn’t seem to be wealthy because he’s a miser through and through. He’s acting like he’s master of the house because, well, why not? It’s an unpleasant situation and he may as well get what small pleasure he can out of it. But Hubert sees the opportunity to pin both the murder and the blackmail on Arthur, so he kills him.
This is really very clever. Both ways of looking at the situation make a lot of sense, and Carr does a good job of keeping both possibilities viable. It could have been the basis for a beautiful reversal, the sort of thing that you find in the best of the Father Brown stories. Instead it’s a mess.
(This is largely due to other aspects to the plot, of course, including the two very stupid impossible crimes I mentioned in the introduction. Even Carr himself isn’t able to muster up much enthusiasm for those. Towards the end, when the twist of the real blackmailer is revealed, he has H.M. say: “That’s the whole secret of this case; and as far as I’m concerned, its the only novelty.”)
This is many people’s least favourite Carr, and it’s easy to see why. Even without considering the “admitted fact” trick, it’s not a very interesting or entertaining book. With the trick, I’d say it’s possible to question whether it’s a mystery at all. Carr himself was very forthright on the topic of fairness in murder mysteries. To cheat so brazenly is a severe breach of trust.
Even people who defend the trick usually admit that the elegant symmetry of the plot is undermined by that first paragraph. For most readers, turning the situation on its head won’t come as a jaw-dropping surprise, but a bemused (or indignant!) “what? But I thought you said at the beginning…?”
It doesn’t even help conceal the important parts of the solution. Hubert is obviously the murderer for any number of other reasons (not least because he was the only suspect not in the room at the time of the murder. In an impossible crime story, where no-one seems to have had the opportunity, anyone with an extra alibi is always suspicious.)
The fact that the motive is concealed doesn’t matter. In a John Dickson Carr mystery, the motive is usually the least of people’s concerns. They’re interested in how it was done, with even the question of who coming quite a way behind.
But is it unfair? Let’s work from the ground up. I don’t think it’s contentious to suggest that if the author explicitly states that “X isn’t the murderer”, and then later says “Ha-ha! X was the murderer after all. I tricked you!”, then that’s a long way from the spirit of the genre. It would certainly be a trick, but not a fair one. If there’s one thing the genre isn’t compatible with, it’s lies. Even the so-called “unreliable narrators” tend to tell a version of the truth at all times (when they do lie, it’s never about the important stuff). Third person narration just can’t be littered with falsehoods. If it’s not in quotation marks, it has to be true.
Carr’s defence, presumably, would be that he didn’t lie. That an admitted fact, in the legal sense, isn’t necessarily a true fact, it’s simply one which both sides have accepted and which doesn’t have to be proven. But there are two issues with that. First, most people are going to be unfamiliar with the term as Carr uses it. For them “admitted” is going to be taken to mean “accepted” rather than “submitted” (as in “submitted for your approval”). Neither of those quite capture the sense, but it’s the second one that’s closest to Carr’s intention. It’s also the weakest of the two meanings, presenting the statement as an assertion rather than an acknowledged truth. But it’s the stronger meaning that’s likely to be in people’s minds, and Carr should have realised that.
The second issue is the more serious. Carr has a long-established habit of using phrases like the “admitted fact” one to mean “obviously I can’t just come out and say X, because this is a book and I have to maintain a veneer of fiction. But there are some things which can’t be 100% verified in the real world, so for the purposes of this puzzle, X is true.”
This is the heart of the issue, and the point that I think people miss when they discuss whether this particular clue is fair. Objectively, I can see that there are situations where saying something is an “admitted fact” and then revealing that it’s false would be perfectly fair. Most of those situations would be where the “admitted fact” phrase occurs in dialogue or in books with a first-person narrator, but I can imagine plenty of third-person books that could get away with this trick. But none of them could be by John Dickson Carr. Reading any mystery – particularly a John Dickson Carr mystery – comes with a lot of context, and it’s this context which makes the clue unfair. (I’m sure no-one will contest my modest assertion that Seeing is Believing is extremely unlikely to be the first Carr anyone reads, and even less likely to be their first mystery of any sort!)
Carr’s books are not naturalistic, nor are they intended to be. They’re puzzles first and foremost, and little attempt is made to hide the presence of the author. In a much-quoted (and overrated) section from Carr’s masterpiece, The Hollow Man, his other series detective, Dr Fell, begins by happily admitting that he’s “in a detective story, and we don’t fool the reader by pretending that we’re not.” In the solution to Seeing is Believing itself, H.M. remarks that Arthur being the killer is only an “admitted fact” (and it’s clearly marked as a quotation), even though no character in the story has used that phrase to him or anyone else. It’s like H.M. has a direct hotline to the author.
Still, the constant presence of the author is a good fit for Carr. His books are a battle of wits, and a game without a clear opponent is less fun. Even if you’re not in it for the challenge, there are good reasons to be glad of Carr’s obvious presence. His mysteries are more complicated than average, and the reasons why his impossible situations seem impossible sometimes require a lot of explanation and reiteration. This is complicated by the fact that the atmosphere and tension often relies on characters having witnessed something they can’t believe has happened. But since everyone is suspicious to the experienced reader, and the experienced reader is who Carr always has in mind, there has to be a way to eliminate the most boring and plausible solution: that the person who saw the impossibility is just making it up. But there’s often no satisfactory way to do this without direct intervention from the author. Appeals to rationality (“but why would X make it up?”) don’t carry much weight in a genre where most characters are lying about something and where books almost always feature one unexpected reason for doing something unusual and often many more.
Carr regularly uses footnotes, or brazen authorial intrusions into the text, stating that certain facts can be taken on trust. He’ll say that certain characters can be trusted, certain observations are true, or even that certain people are innocent (usually minor characters, ones who he’d rather didn’t have to exist but logically have to, given the setting, such as servants, crewmen or staff). The technique is blunt, but simple, and the reader is likely to be forgiving. Everyone is reading Carr for the mystery rather than his insights into society or the human condition; having the conditions and limitations of the puzzle plainly stated is in everyone’s best interests.
(And it’s not like there aren’t plenty of tricks that can be played here. A favourite of Carr’s is to say that certain facts are “strictly true”. This should always set alarm bells ringing and have you on the look out for unjustified assumptions, however minor.)
Carr even makes this sort of assertion, this time as a genuine aid to the reader, in the very same book. At the end of chapter 5, Frank Sharpless says:
“But you’re just as sure about the other thing. So am I. I’ve got eyes. I’ve got ears. I’ll take my Bible oath. I’ll swear to my dying day, that nobody could have got in here either by the windows or by the door!”
To which Carr adds:
And, as a matter of fact, he was perfectly right.
The intent here is clear. It’s an instruction to the reader to ignore the possibility of an intruder sneaking in unnoticed. (It’s also a perfectly fair way of trying to detract suspicion from Hubert, who was outside at the time of the murder, and who committed it by using the window but without actually coming in. There’s that “strictly true” ethos! Although it doesn’t really work to divert suspicion here, because it’s emphasising for a third time just how impossible it would have been for Hubert to have done it. I think even the most gullible reader will be swivelling their head in Hubert’s direction at this point.)
If you grant the premise that 99% of readers will have prior Carr experience, I don’t see how any of them can be expected to realise the statement in chapter 5 is a genuinely helpful message from the author to the reader but the one in chapter 1 is the exact opposite. Either you can trust your opponent to apply the rules consistently or you can’t. If you can’t, then why play in the first place?
And that’s why Seeing is Believing is an unfair mystery. Not because the “admitted fact” trick is inherently cheating or unworkable – although I certainly won’t be appropriating it any time soon – but because it’s John Dickson Carr who’s doing it. It undermines a convention he’s established between himself and the reader, in this and other books, that we’ll accept these little authorial pronouncements because to present them more subtly would be too time-consuming, or require too much skill and effort, and wouldn’t we all rather be getting on with the business of finding out how the killer walked through walls? But it’s undeniably weak technique, and a concession granted magnanimously by the reader in exchange for a fair and interesting mystery. To take advantage of that is very bad form, especially for an author who puts so much stock in doing right by mystery readers.
So what can we learn from this? Is there much point dissecting a book which has so little going for it in the first place? I think it’s important for mystery writers to remember that their books don’t exist in a vacuum. There have never been any hard and fast rules for fairness in mysteries, even when the Detection Club were at the height of their influence. Readers learn the basics as they read their first mysteries, but then they develop their own expectations for different authors. What might be considered fair game from one author could well be unacceptable from another. Authors need to realise what concessions their readers are making for them, and be careful not to take advantage. Once trust in the fairness of an author’s puzzles is undermined, it’s almost possible to get it back.